To improve the habitat of the manned spacecraft (MS) crews with fresh vitamin greens, relatively small and economical space greenhouses (SG) with LED-based lamps (LED) can be used. In order to reduce the costs of onboard resources when growing plants in SG, the method of adaptive search optimization of the lighting parameters of plants during their growth was used. The minimum specific value of the light-dependent part of the equivalent system mass (ESM) per unit of grown biomass (G) was used as the current optimization criterion for SG. The current value of the criterion at plant age t during period of time Δt can be written as: min G (I, t, Δt) = min{[Cv (k3/k2) + CE (k4/k2)·I]/F(I,t) Δt}. Here F(I,t) - is the current net photosynthesis of the crop; CV,CE – cost factors of consumed by SG electric power unit and unit of volume occupied by the SG, in kg/m2 and kg/kW, respectively; k2-k4 - constants of dimension matching; I – lighting attributes which include combination of parameters (intensity, spectrum, time distribution). If lighting spectrum and time distribution are constant, I will equal to average PPFD. Obviously, the cost coefficients CV, СE can depend both on spacecraft design and on space mission specificity – flight duration, task program, crew number, etc. The CV and CE have been calculated for 3 published scenarios of space missions: for a manned station in low Earth orbit, a planetary mission on a Lunar base and a planetary mission on Mars. Special stand was built for adaptive optimization of crop illumination using the crop net photosynthesis as biofeedback. Chinese cabbage was grown in a phytotron under 24-h lighting per day. Above the crop, an LED lamp was mounted with autonomous regulation of supply currents for LEDs with red (660 nm) and white (color temperature 4000 K) radiation. The crop net photosynthesis was automatically measured every 15 min by the rate of CO2 decrease in the closed phytotron air. The control software and hardware complex calculated the value of the optimality criterion G (t) at the time of measurement and calculated the optimal parameters of the supply currents for red and white LEDs, which were implemented at the next iteration of the search. The optimal lighting dynamics for Chinese cabbage crop in the phytotron, minimizing ESM, was experimentally obtained for the last 7 days of the plants’ growth.