Previous research in visual search has mostly focused on controlled, screen-based settings. Scenarios, such as searching in real-world dark spaces, remain less explored. Here, we investigated the effect of three lighting conditions - illuminated (room fully lit), floodlight (room dark, search with flashlight with broad lighting beam), and spotlight (room dark, search with flashlight with narrow lighting beam) - on visual search accuracy, response times, and subsequent memory for objects in a real-world office environment. In each of the three lighting conditions, 36 participants subsequently searched for 60 everyday target objects (half present, half absent) amongst 30 distractors. Manual responses regarding target presence/absence were indicated via a bluetooth device. After the end of the search blocks, participants completed a recognition test to assess target memory (object had to be searched) and visual memory (object was seen). Results revealed that search accuracy was higher for absent targets than for present targets, regardless of lighting condition. However, response times differed significantly: searches were fastest in the illuminated condition, slower in the floodlight condition, and slowest in the spotlight condition. Target memory was most accurate for distractors, lower for present targets, and lowest for absent targets, with no significant influence of lighting. Visual memory showed the opposite pattern: distractors were remembered worse than both present and absent targets, with no difference between the latters. To conclude, lighting conditions did not affect memory performance and search accuracy, suggesting that peripheral vision constraints in flashlight searches impair search time but do not compromise memory encoding.