This paper deals with a multi-item newsvendor problem subject to a budget c
onstraint on the total value of the replenishment quantities. Fixed costs f
or non-zero replenishments have been explicitly considered. Dynamic program
ming procedures are presented for two situations: (i) where the end item de
mand distributions are assumed known (illustrated for the case of normally
distributed demand) and (ii) a distribution free approach where only the fi
rst two moments of the distributions are assumed known. Ln addition, simple
and efficient heuristic algorithms have been developed. Computational expe
riments show that the performance of the heuristics are excellent based on
a set of test problems.