Single-molecule manipulation has allowed the forced unfolding of multidomain proteins. Here we develop a theory that not only explains these experiments, but also points out a number of difficulties in their interpretation and makes suggestions for further experiments. Our theory is valid for essentially any molecule that can be unfolded in the AFM: as an example we present force-extension curves for the unfolding of both titin and RNA hairpins. For titin we reproduce force-extension curves, the dependence of break force on pulling speed, and break-force distributions, and also validate two common experimental views: unfolding titin Ig domains can be explained as stepwise increases in contour length, and increasing force peaks in native Ig sequences represent a hierarchy of bond strengths.