Recently, a high prevalence of morning-types was reported among trained South African endurance athletes. Proposed explanations for this observation were that either the chronotype of these athletes is better suited to coping with the early-morning start times of endurance events in South Africa; or habitual early waking for training or endurance events may have conditioned the athletes to adapt and become morning-types. The South African endurance athletes also had earlier chronotypes compared to a control population of less active individuals, suggesting that individuals who are more physically active may have earlier chronotypes. However, since both the South African athlete and control groups showed an overrepresentation of morning-types compared to European and American populations, the South African climate may in part have explained this bias towards morningness. Given the latitude and climate differences between South Africa and the Netherlands, and that South African marathons typically start at about 06:30 while those in the Netherlands start later (±11:00), comparison of South African and Dutch marathon runners and active controls would allow for simultaneous assessment of the effects of marathon start time, degree of physical activity and climate on chronotype. Therefore, the primary aims of this study were: (i) to assess the effect of marathon start time on chronotype in marathon runners and (ii) to determine the extent to which either degree of physical activity or climate might explain the bias towards morningness observed in South African athletes and controls. A secondary aim was to determine whether any relationships exist between chronotype, PERIOD3 (PER3) variable number tandem repeat (VNTR) polymorphism genotype, habitual training habits and marathon performance. Trained male marathon runners from South Africa (n = 95) and the Netherlands (n = 90), and active but non-competitive male controls from South Africa (n = 97) and the Netherlands (n = 98) completed a questionnaire capturing demographics, training and race history, as well as the Horne–Östberg morningness–eveningness personality questionnaire. All participants donated buccal cell samples from which genomic DNA was extracted and polymerase chain reaction analysis was used to genotype them for the PER3 VNTR polymorphism, which has previously been associated with chronotype. The main finding was that South African runners were significantly more morning-orientated than Dutch runners suggesting that participation in an endurance sport with an earlier start time may influence chronotype. Secondly, both the South African and Dutch runners were significantly more morning-orientated than their respective control groups, indicating that individuals who train for and participate in recreational endurance sport races have an earlier chronotype than physically active but non-competitive males. Thirdly, mean chronotype scores were similar between the South African and Dutch control groups, suggesting that climate does not seem to affect chronotype in these groups. Fourthly, the PER3 VNTR polymorphism distribution was similar between the four groups and was not associated with chronotype, suggesting that the difference in chronotype between the four groups in this study is not explained by the PER3 VNTR genotype. Lastly, in the South African runners group, a higher preference for mornings was associated with a better personal best half-marathon and current marathon performance.