The objective of this study was to quantify the progress towards yield improvement in maize (Zea mays L.) using phenotypic mass selection for superior attributes including earlier maturity, increased ear length, 1000 grain weight and grain yield. One cycle of phenotypic mass selection using multiple trait selection criteria was carried out in maize variety, Dehqan. The experiment was carried out in two growing seasons during the year 2001 and 2002, raising one crop per year. In Kharif season 2001 the source population Dehqan was sown on an area of 500 m2 and selection for the desired attributes was made using phenotypic mass selection. The selected (C1 and the original population (C0) were grown in the next crop season (2002) along with two checks i.e. a maize hybrid, Cargill 919 and an open pollinated maize variety Sarhad Yellow. The four maize populations were evaluated in a randomized complete block design (RCBD) using four replications with six rows plot-1, each 5 m long and 0.75 m apart. The results showed that although phenotypic superiority of the selected material over the original population was obvious, however, differences between the two populations were statistically non significant for important ear and grain traits. The selected population manifested higher ear length (25.25 cm), higher grain moisture at harvest, increased fresh ear weight (4.49 kg plot-1) and grain yield (1527.28 kg ha-1) when compared with the original population. This study indicated that one cycle of phenotypic mass selection helped to some extent in accumulating favorable alleles in the selected version of maize variety Dehqan. A more rational basis for utilization of plant traits in a breeding program would appear to involve the simultaneous use of several traits.