Purpose: Chronic nonspecific low back pain is one of the most demanding public health problems regarding treatment and rehabilitation now days. Low back pain is usually defined as pain, muscles’ tension, discomfort and/or stiffness caudally from the rib arch and cranially from the lower gluteal fold with pain irradiation to the leg (ischialgia) or without it. In majority of patients, there is no any clear cause of a low back pain, so even 85% of patients do not receive a specific causal diagnosis. The low back pain could be significantly incapacitating and considerably reduce the quality of life of the individual. It affects various domains of life – physical, psychical, even social relationships and therefore it is a typical biopsychosomatic disease. Even 65-85% of adults experience at least one episode of low back pain during his/her lifetime. There are numerous heterogeneous risk factors for low back pain. They include demographic, physical, social-economic, psychologic and work related risk factors. According to a simple, practical and generally accepted classification the low back pain may be classified as nonspecific and specific low back pain and low back pain where nerve roots are affected, i.e radiculopathy. In nonspecific low back pain we cannot find a specific cause and it is assumed that is mechanical low back pain. The term specific low back pain is used for any localized source of pain because of change of spine structure which could result in pain and/or disablement, whether it is a matter of destructive disease (tumor, infection) or large neurologic deficit (spinal stenosis, discus hernia). There are many generally accepted methods for measuring pain intensity, but in practice is mostly used visual analog scale (VAS). For estimation of function and quality of life there are used specific disease questionnaires and for chronic low back pain mostly used are Oswestry Low Back Pain Disability Index (ODI) and Rolland-Morris Low Back Pain Measure (RM). Because of heterogeneity of risk factors and complex pathophysiological source of pain, treatment of patients with low back pain is complex. It includes various pharmacological, non-pharmacological measures, psychological and social intervention that demands interdisciplinary approach. As pharmacotherapy is concerned, most frequently prescribed drugs are nonsteroid anti-inflammatory drugs, then paracetamol and weak opioids. Antidepressants, anticonvulsants, muscles relaxants and topic analgesics are less frequently prescribed. Drugs of choice in patients with nonspecific chronic low back pain are nonsteroid anti-inflammatory drugs – diclofenac, ibuprofen, ketoprofen, piroxicam, celecoxib etc. No one of them is preferred, but when prescribed, one should be aware of its side effects, above all those gastrointestinal and cardiovascular. The basis of non-pharmacologic treatment of patients with nonspecific low back pain is kinesitherapy. There is a wide range of evidence that therapeutic exercises have positive preventive and curative impact on pain decrease and/or improvement of function. According to European guidelines for treatment of patients with chronic low back pain, there is moderate evidence that therapeutic exercises are more effective than other passive procedures, not preferring any type of exercises. Through numerous original researches, meta analyses and systematic reviews the effect of various types of exercises regarding pain reduction and improvement of function in patients with chronic nonspecific low back pain had been analyzed. It was concluded that neither one of the exercises is superior to the other, also there is sufficient information on intensity, frequency and load of exercises, so it wasn’t possible to define the optimum dose of exercising. Also, the effect of therapeutic exercises in patients with low back pain is mostly small to moderate which could be explained by non-conformity between heterogeneousness of problems and uniform approach to therapeutic exercises. Available literature does not support convincing connection between changes in clinical outcome and changes in body function after exercise therapy. It is assumed that positive effects of exercising by patients with chronic low back pain is more “central” than local, which perhaps includes psychologic, cognitive and/or neurophysiologic (cortical) adaptation. Throughout last few decades, it has been emphasized that changes observed in patients with chronic nonspecific low back pain are not found only in local musculoskeletal structures of a lumbar spine, while these patients also have morphologic and biochemical changes in certain centers of the central nervous system. Specific exercise program that includes motor learning (combination of contralateral movements, stimulation of memory and attention), uses mechanic, electric stimuli, stimuli of spatial orientation and balance (proprioception) stimulates a part of central nervous system, that participates in creating and experiencing the sense of pain, however also locally strengthens musculoskeletal structures in the area of pain. The aim of this research was to prove that specific therapeutic program of exercises, including sensorimotor exercise, exercises of proprioception and postural control, in combination with usage of external stimulus (sensory, mechanical, electric) to the area of lumbosacral spine during exercising, is more effective than conventional therapeutic program of exercises used in rehabilitation in patients with chronic nonspecific low back pain. Methods: This research included 88 patients of both gender, aged 25 to 70, who consecutively visited physiatrists practice of the Clinic for Rheumatology, Physical Medicine and Rehabilitation at Clinical Hospital Center Sestre Milosrdnice in Zagreb. During the time of research, 8 patients withdrew (10%). Criteria for enrolment were as follows: patients with localized low back pain with or without irradiation of pain to a leg to the knee, in duration of at least 3 months with pain intensity exceeding 30 mm measured on 100 mm visual analogous scale (VAS), without “red flags” that indicates serious pathology. Exclusion criteria were: malignant tumors diagnosed in last 5 years (except non-melanoma skin cancer), non-infectious, and infectious inflammation (acute or chronic), including inflammatory rheumatic diseases, increased body temperature of any etiology, serious form of metabolic diseases (diabetes, thyroid gland disease etc.), more difficult diseases of cardiovascular system, difficult neurologic diseases and states (e.g. sclerosis multiple, cerebrovascular insult), serious psychical diseases and states, fresh trauma of lumbar spine (3 months past from interrogation), heart rate electrostimulation device, pregnancy and patients in whom in the last 3 months physical therapy was applied in the area of lumbar region. Patients were divided into two groups with the same number of individuals, without formal randomization. However, equal distribution of patients by age, gender and intensity of pain in each of two groups was taken into consideration. First group of 40 patients performed conventional therapeutic program of strength exercises (flexion extension) and stretching exercises for lumbosacral spine. The other group of 40 patients performed specific therapeutic program of sensorimotor exercises, exercises of proprioception and postural balance. Conventional therapeutic program consisted of 7 static and 11 dynamic strengthening exercises and 1 stretching exercise. Each static strengthening exercise was performed for 5 seconds, upon which the minimum of 5 seconds relax in introduction set of 5 repetitions. One repetition was added every day, reaching the maximum of 10 repetitions. Each dynamic strengthening exercise was performed initially 5 times, increasing every day for one repetition, up to 10 repetitions in one set. The following principle was implemented: patient had to perform the movement while breathing in through the nose, and relax the muscle while breathing out through the mouth. The set of exercises ended with maximum stretching of the whole spine. Stretching lasted for 5 to 10 seconds relax also 5 to 10 seconds and exercise was repeated 5 to 10 times. Specific therapeutic program of exercises (sensorimotor exercises) included exercises of motor control (spinal segmental stabilization exercises or stabilization exercises), exercises of postural control (balance and coordination exercises) and exercises of proprioception to which different type of stimulus in the area of lumbosacral spine were added during the exercises. Stimulus being used was in form of warmth stimulus – warm paraphine pack applied during 3 exercises in supine position and cold pack applied during 2 exercises in supine position with additional load of ½ kg and use of easily compressive, medium large ball. After that, mechanic nonnociceptive stimulus in form of spikey wooden roller was used in 2 exercises, with additional thera band load (extra strong) and medicine ball. During 3 exercises stimulus was induced by electric current, i.e. transcutaneous neuromuscular electric stimulation in pronated position of the examinee. In one exercise kinesiology ball was used and in other one Swedish ladder, while two exercises balance board. Specific therapeutic program of exercises was designed in a way that forced candidates to master demanding tasks, i.e. exercises of coordinating laterolateral distal segments of body with stabilization of a torso, regular breathing and awareness of movements. During the performance of sensorimotor exercises, different types of before described exercises were also introduced and combined, whose basis was postural and motor control, resulting from complex interactions of sensoric and musculoskeletal system, integrated and modified within the central nervous system in response to a change of state of the environment. Evaluation was made by a questionnaire, in which we registered variables directly before the beginning of intervention: age (years), sex (male/female), body height (centimeters), body mass (kg), body mass index (height in meters2/kilograms), profession, years of employment (years), physical work (light, moderate, heavy, very heavy), duration of a pain in lumbar spine (months), analgesics (type, daily dose, duration). Co-primary outcomes were: pain in lumbar spine in rest (laying) (measured on 100 mm of horizontal visual analogous scale (VAS) within the last 24 hours), pain in lumbar spine in movement (VAS) and degree of disablement measured by the Oswestry Disability Index (Chapman et al., 2011). Other observed outcomes were: global patient’s and doctor’s evaluation (on 100 mm VAS), range of spine movement (inclination, reclination, lateroflexion –measured in cm) and tension of paravertebral lumbar muscles (0, 1, 2, 3). These variables were measured directly before and after the treatment and one month after the treatment. For the purpose of objective evaluation of results, the type and dose of drugs should not be changed during the research. In case of increase pain, the patients were allowed taking paracetamol, with registering a daily dose. In accordance with the aim of research, the applied data processing referred to establishment of central and dispersive indicators of all measured variables, establishment of significance of differences in measured variables among groups of examinees of sensorimotor and conventional group at initial testing, after 20 treatments and 30 days after the treatment and establishment of significance of treatment effects in measured variable in examinees of sensorimotor and conventional group. With non-parametric Mann-Whitney there was verified significance of differences in measured variables among groups of examinees of sensorimotor and conventional group at initial testing, after 20 treatments and 30 days after performed treatments. With Friedman’s test with Wilcoxon match-paired Signed-Rang post hoc test there was verified significance of treatment effects in measured variables in examinees of sensorimotor group and conventional group. Data were processed by the computer program SPSS for operative system Windows, version 23.0. P- value <0.05 was statistically significant. Results: In group of patients who performed sensorimotor exercises it was showed statistically significant improvement in variables: pain in rest and movement, global patient’s and examiner’s health assessment and total sum of the Oswestry disability questionnaire as in certain variables of the range of movement in lumbar spine, in relation to the same variables in examinees who performed conventional therapeutic program of exercises immediately after the therapy and one month after the therapy. The effect of improvement of variables pain in rest and movement, global patient’s and examiner’s health assessment between initial measurement and measurement immediately after the therapy and initial measurement and one month upon completed the therapy was strong (r>0.5). It was also strong effect of exercises on improvement of range of movement (Schober measure, reclination and lateral flexion) and at evaluation of performing everyday life functions by Oswestry disability questionnaire. In evaluating single variables of Oswestry questionnaire we concluded that exercises performed by patients of sensorimotor group had medium to strong positive influence on all variables, including the easiest activity (walking) and the heaviest activity (standing up), contrary to positive effect of conventional exercises on 7 to 10 variables of the Oswestry disability questionnaire which was medium (if we exclude the effect on pain intensity (IB) which was medium strong to strong). As far as single variables referring to pain, global patient’s and examiner’s health assessment and spine functions are concerned, the effect of conventional exercises was positive in all variables, but smaller in relation to sensorimotor exercises, i.e. it was medium to strong. Long-term reduction of pain and statistically more significant reduction of pain in variables pain in rest and movement recorded in sensorimotor group could be attributed to the effect of specific training to central nervous system. With reduction of pain intensity it was also observed the improvement of general health status, because pain is, although subjective, important factor of quality of life that could be observed through global health assessment as well. By regression of pain, tension of paravertebral muscles was also reduced, which is objective indication of pain intensity and dysfunction of the low back part. By regression of pain and positive effect of exercising on local musculoskeletal structures, there has been an improvement in variables that evaluate function of a spine (inclination, reclination and lateral flexion) and consequently the improvement of functional ability. Discussion and conclusion: In this research, it was confirmed that conventional therapeutic program of flexion extension exercises of strengthening lumbosacral muscles as well as specific therapeutic program of sensorimotor exercises reduce pain and improve functional ability of patients with chronic nonspecific low back pain. Nowadays, it is widely adopted that patients with chronic nonspecific low back pain face cortical reorganization and central nervous system changes. Brain changes, visible in primary and secondary somatosensory cortex, but also in primary motor cortex result in sensory (tactile precision), perceptive (changed pictures of body) and motor difficulties (motor control). These neurophysiological changes (particularly in primary somatosensory cortex) correlate with intensity and duration of pain. Patients with chronic nonspecific low back pain have proven morphologic changes of the central nervous system in the gray matter area (insula, temporal lobe, primary somatosensory lobe – S1, dorsolateral prefrontal cortex), increased brain activity in medial prefrontal cortex, cingulate cortex, amygdala, insula, integrational sensorimotor regions and also disorder of neural network of a central nervous system. These conclusions were evidenced through functional magnetic resonance (MR). Furthermore; there is an increased activity in areas responsible for pain (S1, secondary somatosensory lobe (S2), rear cingular cortex and insula) and reduced activity in patients in periaqueductal gray matter upon mechanic stimulus in comparison to healthy individuals. Visible morphologic changes are also present in white matter (corpus callosum, above corpus callosum and in internal capsule). Different responses to thermal stimulation in areas of medial prefrontal cortex and insula, as well as different responses to electrical stimulation in the rear cingular lobus are observed between patients with chronic nonspecific low back pain and healthy individuals. Above mentioned areas of the brain, in which changes are observed in patients with chronic nonspecific low back pain refer to areas responsible for pain (insula participates in processing information for pain and temperature), integration of visual, auditive, sensory speech information and emotions (temporal lobe), spatial orientation, proprioception, processing somatic sensory stimulus (primary somatosensory cortex (S1), retention and real time processing information – “working memory” (dorsolateral prefrontal cortex). White matter damage may disturb connection of brain hemispheres and cause difficulties in performing actions that require mutual coordination of two or more brain regions. These particular neurophysiologic alterations connected to changes in structure, function or organization of neurologic system occur during whole lifetime. They are being affected by internal stimulus as e.g. cognitive processes, internal biochemical and morphologic changes, reaction of the central nervous system to sensory stimulus and external stimulus like motor learning and peripheral sensory stimulation. Before mentioned damages of nervous system could be the cause of chronic pain and relapsing pain in patients with chronic nonspecific low back pain. Neuroplasticity of a brain provides possibility of discovering and creating new experiences, developing new behavioral patterns and creating therapeutic processes that could affect the neurologic structure changes. Such a therapy includes e.g. training of sensory discrimination, peripheral electric stimulation or motor learning. Motor learning requires focused attention, increase of complexity of a task and excellent interaction with the therapist, what can result in pain modulating. During the training, it is advised to use also the principle of laterality and introduction of mirrors. Motor learning exercises may stimulate cortical reorganization in primary motor cortex, regulate excessive corticospinal stimulation connected with tension of surface-paravertebral muscles and regulate the lack of intracortical inhibition necessary for motor planning. Patients with chronic, nonspecific low back pain have deranged scheme of body and extremities, reduced lumbosacral sensitivity and proprioceptive sharpness, difficulties in performing sensorimotor task and deficiency of postural control. Moreover, patients with chronic nonspecific low back pain show increased activations of brain during the non-nociceptive external stimulus (e.g. pressure), as well as body induced stimulus (e.g. performing motor task) in areas responsible for pain. Taking into consideration above mentioned observations it is necessary to conduct global therapeutic approach which is focused to influence central nervous factors on regression of pain and improvement of disablement in patients with chronic nonspecific low back pain. Patients with chronic pain under the influence of a long nociceptive stimulation experience changes in morphology and biochemistry of brain and relation of neural networks. These changes mostly take place in brain centers responsible for sensory, affective and cognitive dimension of pain. Superior effect of specific therapeutic exercise program in comparison with conventional therapeutic exercise program may be attributed to practical application of recent knowledges based on scientific evidences, using mechanisms of treatment on changes in local muscles and surrounding musculoskeletal structures in the area of lumbosacral spine, as well to effects of specific exercises, above all motor control exercises with various kinds of stimuli on brain centers, primarily on primary motor and somatosensory brain cortex.