Deployment of resistant varieties is one major approach to controlling cassava bacterial blight (CBB), caused by Xanthomonas axonopodis pv. manihotis (Xam). To understand the genetic determinism of resistance to CBB, the use of reliable parameters measuring resistance is necessary. In order to test a relevant method for evaluation of quantitative resistance for mapping QTL (quantitative trait loci), the response of 150 F1 individuals, inoculated with four different Xam strains (CIO-84, CIO-1, CIO-136 and CIO-295), was assessed under controlled conditions.We used two types of evaluations at different intervals after inoculation, one based on a scale of 0 to 5 and the second based on the determination of the bacterial population in the vascular system. Both evaluation types revealed interaction between strains and F1 genotypes. Population values at 3 and 6 cm from the point of inoculation showed a high level of correlation. By performing an association analysis, at 7 and 15 days after inoculation, a significant positive correlation between both evaluation types was obtained. However, the disease rating at 30 days did not correlate with bacterial populations at either 7 or 15 days after inoculation, except for one strain, CIO-84. Evaluation of the bacterial population in stem tissues is time and labour consuming, consequently, for a rapid and reliable assessment of CBB resistance for QTL analysis, we strongly recommend evaluation based on the use of a symptom scale.