Long-day photoperiod is known to positively affect milk production in confinement dairy systems, and it has been hypothesized that pineal melatonin (MT) secretion plays a substantial role in this process. Specialized mammalian photoreceptors that regulate MT secretion are optimally stimulated by short wavelength blue light. We investigated the blue light intensity administered to one eye required to suppress MT secretion in nonlactating dairy cows, and subsequently examined effects on milk production in grazing dairy cows. Following a 14-d light-dark 8:16 h environmental conditioning period, 5 nonlactating Holstein-Friesian cows were exposed to treatments of <1, 70, 125, 175, and 225 lx for 8 additional hours using a 5 × 5 Latin square design. Light was administered via headpieces fitted with light-emitting diodes emitting blue light (465 nm) to the right eye. All cows were then exposed to a light-dark 16:8 h cycle for one night via the indoor lighting system (>200 lx white light). Plasma samples collected at regular intervals were assayed for MT. A dose-dependent effect of light treatment on mean circulating MT concentrations (and 95% CI) was observed [9.4 (7.2, 12.3), 5.0 (3.8, 6.6), 4.4 (3.3, 5.7), 3.3 (2.5, 4.3) and 1.7 (1.3, 2.3) pg/mL for treatments of 0, 70, 125, 175, and 225 lx, respectively. Only the 225 lx treatment acutely suppressed plasma melatonin concentration to levels similar to the light-dark 16:8 h treatment [1.9 (1.4, 2.5) pg/mL]. Forty spring-calving cows were blocked on parity, calving date and Economic Breeding Index for milk production and assigned to the control treatment or blue light to a single eye (LT) treatment from calving through 32 wk of lactation. The cows assigned to LT treatment were fitted with headpieces providing 225 lx of blue light to the right eye from 1700 until 0000 h. Mean milk production (and 95% CI) during 32 wk of lactation was not affected by treatment [20.3 (19.3, 21.3) vs. 20.9 (19.8, 22.0) kg/d, control and LT, respectively]. Within multiparous cows, a treatment by week interaction was detected, whereby LT treatment increased milk production during the first 12 wk of lactation [25.8 (24.3, 27.3) vs. 28.0 (26.5, 29.5) kg/d; +8.5%], but had no effect thereafter. Treatment did not affect plasma insulin-like growth factor 1. We identified the blue light intensity to one eye required to acutely suppress MT concentrations. Transient favorable effects on milk production were observed in multiparous cows. It remains unclear how single-eye blue light treatment affects galactopoiesis in grazing dairy cows, and further research is needed to explore whether this modality of light delivery represents a useful means to aid productivity in pasture-based dairy systems.