Five peach cultivars [Prunus persica (L.) Batch] with different maturity dates were subjected to sink–source manipulation by girdling to isolate 1-year-old fruit-bearing shoots. Four treatments were performed: fruit were removed (−fruit); one fruit (+1 fruit) and two fruit (+2 fruit) were kept inside two girdling cuts; and two fruit were kept outside two girdling cuts (−fruit*). Photosynthetic responses for the five cultivars were similar and did not show genotypic differences. Generally, net photosynthetic rate (Pn), stomatal conductance (gs), and transpiration rate (E) were higher, and leaf temperature (Tl) was lower in +2 fruit than in +1 fruit, followed by −fruit and −fruit* which were not different. The results also indicated that water outflow from fruit into leaves did not influence photosynthesis, and lower photosynthesis in −fruit treatment was not due to water status of source leaves influenced by removing fruit. Pn tended to increase with Tl until Tl reached a critical level. Beyond the critical temperature level, Pn generally decreased. The critical Tl was roughly identified as 34–37 °C for the five cultivars. Both higher and lower substomatal CO2 (Ci) levels occurred in −fruit and −fruit* treatments than in +1 fruit and +2 fruit treatments, indicating that decreased Pn could be due to both nonstomatal and stomatal limitations. Further analysis of the relationship between Ci and photosynthetically active radiation (PAR) showed that nonstomatal limitation under low sink demand took place mostly under high PAR. Thus, high light intensity, combined with Tl may play an important role in leaf photosynthetic regulation.