If a set of equations $E\!\cup \!Ax$ is such that $E$ is confluent,
terminating, and coherent modulo $Ax$, narrowing with $E$ modulo $Ax$
provides a complete $E\!\cup \!Ax$-unification algorithm.  However,
except for the hopelessly inefficient case of full narrowing, nothing
seems to be known about effective narrowing strategies in the general
modulo case beyond the quite depressing observation that basic
narrowing is \emph{incomplete} modulo $AC$.  In this work we propose
an effective strategy based on the idea of the $E\!\cup \!Ax$-variants
of a term that we call \emph{folding variant narrowing}.  This
strategy is \emph{complete}, both for computing $E\!\cup
\!Ax$-unifiers and for computing a minimal complete set of variants
for any input term.  And it is \emph{optimally variant terminating} in
the sense of terminating for an input term $t$ iff $t$ has a finite,
complete set of variants.  The applications of folding variant
narrowing go beyond providing a complete $E\!\cup \!Ax$-unification
algorithm: computing the $E\!\cup \!Ax$-variants of a term may be just
as important as computing $E\!\cup \!Ax$-unifiers in recent
applications of folding variant narrowing such as termination methods
modulo axioms, and checking confluence and coherence of rules modulo
axioms.