Seasonal reproduction is a common feature among various mammalian species of the temperate latitudes. Ewes are well known for seasonal breeding activity which usually appears at the end of summer or at the beginning of autumn and finish in late winter or at the very beginning of spring. The seasonality of reproductive activity in sheep breeds in temperate latitudes is controlled by the photoperiod. In sheep, short photoperiods stimulate pineal gland to synthesize hormone melatonin, while long photoperiods inhibit pineal gland in melanin synthesis. Therefore, melatonin inform animal on photoperiod changes and has stimulating effect on hypothalamus which release Gonadotropinreleasing hormone. Pituitary gland stimulated by hypothalamus release Luteinizing hormone which affects the growth of oocytes. In sheep melatonin exerts its function through specific MT1 G-protein receptor. Gene for this receptor (MTNR1A gene) is located on chromosome 26 and has two polymorphic sites at position 606 and 612 which are associated to aseasonal breeding activity. Considering that MTNR1A gene has been proposed as candidate gene for out of season breeding in sheep, present study evaluates the effects of polymorphisms at positions 606 and 612 on sheep seasonal reproduction. The objectives of the study were: 1) determine the frequency of genotypes and allele variants of MTNR1A gene at positions 606 and 612 (reference sequence U14109), 2) determine the presence of polymorphisms on the studied sequences on the second exon of MTNR1A gene, 3) determine the effect of MTNR1A gene polymorphisms on seasonal reproduction, age at first lambing and interval between lambing. Depending on breeding place the biological material (blood, tissue, hair) and sheep reproductive parameters were collected during 2015 and 2016 year in the area of Republic of Croatia and Republic of Slovenia. Samples of Croatian indigenous sheep breeds (Tsigai sheep, Dalmatian Pramenka, Dubrovnik sheep, Istrian sheep, Lika Pramenka sheep and Pag Island sheep), imported noble sheep breeds (East Friesian sheep, Romanov sheep and Suffolk), as well as mouflon (Ovis musimon) samples (from hunting grounds) were collected in Croatia. In Slovenia were collected samples of Slovenian indigenous breed: Bovec sheep. Except for noble sheep breeds which are characterized by a prolonged period of mating season, the rest of breeds express seasonal sexual activity and they are considered as seasonal polyestric sheep breeds. From 439 collected samples genomic DNA was extracted using GenElute® Mammalian Genomic DNA Miniprep Kit (Sigma-Aldrich). After DNA extraction, the main part of second exon MTNR1A gene was amplified using specific primers (forward: 5'-TGTGTTTGTGGTGAGCCTGG-3'; reverse: 3'ATTTGCGTTTGGGAGAGGTA-5') adopted from Messer et al (1997). Obtained sequence length was 824 base pairs which was then subjected to enzymatic resolution with RsaI and MnlI restriction endonucleases. Polymorphism at position 606 was detected using RsaI, while polymorphism at position 612 was detected using MnlI restriction endonucleases. Samples were genotyped as follows: CC, CT or TT on locus 606 and GG, GA or AA on the locus 612 of MTNR1A gene. In five samples of each genotype determined with MnlI restriction endonucleases nucleotide sequence were determined in the target sequence of investigated DNA and confirmed the presence of polymorphisms in the studied loci (606 and 612). Reproductive parameters were date of birth of studied animal and dates of all her lambing records. Season of birth and seasons of lambings were calculated on the basis of date of birth and dates of lambings. Date of birth and date of first lambing were used to determine length of interval (in days) required for animal to have first offspring. Dates of subsequent lambings were used to calculate the average interval between lambing. For genotype effects estimation on first lambing on loci 606 and 612 the following linear model was used: Y1ijklm = µ + tri + srj + pk + Gl + eijklm. For genotype effects estimation on interval between lambing on loci 606 and 612 the following linear model was used: Y2ijklmn = β0 + β1xijklmn + tri + srj + pk + Gl + eijklmn. Considering the length of restriction fragments observer on agarose gel under UV light all three genotypes were detected at locus 606 and 612 of the MTNR1A gene in studied populations. Genotype frequencies at locus 606 were: 0.26, 0.45 and 0.29 for CC, CT and TT, respectively. Genotype frequencies at locus 612 were: 0.46, 0.39 and 0.15 for GG, GA and AA, respectively. Allele frequencies were: 0.49 for C, 0.51 for T, 0.66 for G and 0.34 for A allele. Differences in determined genotype frequencies between mouflon and all studied sheep breeds were statistically significant at both loci (P<0.001; P<0.05). Also statistically significant differences in determined genotypes frequencies were found between investigated sheep breeds. High frequency of the allele C on locus 606 was determined in mouflon, Romanov and Suffolk breeds while seasonal polyestric sheep breeds had higher frequencies of allele T. The locus 612 in almost all studied sheep breeds revealed a high frequency of the allele G (0.50 to 0.95) except mouflon in which we identified a high frequency of allele A (0.93). The most frequent combinations of genotypes at position 606 and 612 were: CTGA (27.79%), TTGG (27.11% and CTGG (16.86%). The examined populations were in Hardy-Weinberg equilibrium for both loci, except for locus 612 in Bovec sheep (P<0.01) and on both loci (606 and 612) in Dubrovnik sheep (P<0.01). The polymorphism at locus 606 was expressed by the presence of cytosine (C allele) instead of thymine (T allele), while polymorphisms at loci 612 was expressed by the presence of guanine (G allele) instead of adenine (A allele). Except those two mutations another six mutation were detected at the investigated DNA sequence. Those six mutations were: G453T, G706A, G783, G801, C891T and C893A. The mutation in position G706A and C893A led to an amino acid change in position 220 and 282, while all other mutations were silent. Sheep lambing were recorded in all four seasons, and regardless of genotype the highest number of lambing was recorded in winter (61.89%). Although out of season (summer and fall) was recorded only 11% of lambing, ewes with homozygous (CC/606; GG/612) and heterozygous (CT/606; GA/612) genotype dominated. In spring lambing dominated genotype combination CTGA (32.37%), in summer genotypes CCGG, CTGG CTGA with 25% each, in fall genotypes TTGG and CTGA and in winter genotypes CTGA and TTGG. Age of first lambing regarding genotype at position 606 was very uniform from 513.8 days in genotype CC to 519 days in genotype TT. On the other side at position 612 animals with genotype AA needed 434.9 days to first lambing, animals with GG genotype 525.1 days and GA genotype 521.6 days. Differences between genotypes on both loci were not statistically significant. Combination CTAA had the most favourable effect on lower age at first lambing. Individuals with mentioned genotype lambed with average age of 375.5 days while individuals with CCGG genotype were the oldest at first lambing with 587.9 days. Average interval between lambing in observed genotypes were as follow: 360.4, 358.5 and 361.3 days in CC, CT and TT genotype and 359.3, 360.7 and 356.6 days in GG, GA and AA genotype, respectively. Differences between genotypes on both loci were not statistically significant. Combination of genotypes TTAA had the smallest interval between lambing of 352.3 days while genotype combination TTGA had the biggest lambing interval (394.1 days). In the end we can conclude that polymorphisms at loci 606 and 612 of MTNR1A gene were observed in studied populations. Despite the absence of statistically significant effect of genotypes at loci 606 and 612 on studied parameters some positive effect of MTNR1A gene polymorphisms in some sheep breeds were observed.