The purpose of this study was to refine scales to measure Health Belie
f Model concepts of benefits and barriers using the context of mammogr
aphy screening. Both scales used a summated Likert format with five re
sponse options. Items were first subjected to content analysis by nati
onal experts. Data were then collected from a random sample of 581 wom
en who were 35 years or over. Retest data were gathered on a subset of
women similar to 4-6 weeks later. Construct validity was established
using confirmatory factor analysis and exploratory factor analysis wit
h results confirming independence of constructs. Construct validity wa
s also tested by developing hypotheses based on theoretical relationsh
ips. Cronbach alpha reliability coefficients for the benefits scale wa
s 0.79 and for the barriers scale was 0.73. Suggestions for further re
visions are offered.