Several ideas, based upon published information on rubber and other crops, are discussed in relation to rubber research. First, a simple, formal theory of rubber yield as an aspect of growth and partition is given. It uses three parameters and could be applied, it is suggested, to interpretation of the breeding history of the crop. Secondly, relation between tapping cost and yield, commonly described as negatively linear, is shown to be hyperbolic and the difference is of some consequence near the limits. Thirdly, the predictive capacity of variety trials is then discussed; this has rarely been tested, or even questioned. In rubber, agreement is fairly good but the regression of commercial yields on trials yields over varieties diverges from unity. Finally, vertical and horizontal resistances to airborne fungal diseases are briefly reviewed. The former (roughly equivalent to major-gene resistance) is unreliable and the latter much to be preferred. The implications for breeding resistance to SALB (South American Leaf Blight) in rubber are clear; some horizontal resistance to the disease is known to be present.