In this study, sample unit size and number of samples for major insect-pests of Safflower (Carthamus tinctorius L.) were estimated during 2015-2016. The experiments were laid out in randomized complete block design with three replications in which the experimental area measuring 2000 m2 was divided into 45 subplots. Four samples of different unit sizes were taken as one plant in an area of 0.12m2, two plants in an area of 0.25 m2, four plants in an area of 0.50 m2, and eight plants in an area of 1.00 m2. Based on 2- years data, the sample unit size of two plants for estimating the population of Acanthiophilus helianthi, four plants for the population of Helicoverpa peltigera, Lygus pratensis, and Larinus flavescens as well as eight plants for estimating the population of Oxycarnus pallens were found to be accurate on safflower. Likewise, the larger numbers of samples were calculated to estimate the population of major insect-pest i.e.75 per 0.2 ha though was not significantly higher to fewer numbers (45 and 15 per 0.2 ha) when population means were compared but the precision values suggested that accuracy increased when the fewer number of samples were taken and cost of sampling decreased. Therefore, 45 samples were sufficient to estimate the population of major insect-pests in safflower.

Keywords: Safflower, major insect-pests sampling, sample unit size, population estimates