Let $W(G)$ be the Wiener index of a graph $G$. We say that a vertex $v \in V(G)$ is a Šoltés vertex in $G$ if $W(G - v) = W(G)$, i.e. the Wiener index does not change if the vertex $v$ is removed. In 1991, Šoltés posed the problem of identifying all connected graphs $G$ with the property that all vertices of $G$ are Šoltés vertices. The only such graph known to this day is $C_{11}$. As the original problem appears to be too challenging, several relaxations were studied: one may look for graphs with at least $k$ Šoltés vertices; or one may look for $α$-Šoltés graphs, i.e. graphs where the ratio between the number of Šoltés vertices and the order of the graph is at least $α$. Note that the original problem is, in fact, to find all $1$-Šoltés graphs. We intuitively believe that every $1$-Šoltés graph has to be regular and has to possess a high degree of symmetry. Therefore, we are interested in regular graphs that contain one or more Šoltés vertices. In this paper, we present several partial results. For every $r\ge 1$ we describe a construction of an infinite family of cubic $2$-connected graphs with at least $2^r$ Šoltés vertices. Moreover, we report that a computer search on publicly available collections of vertex-transitive graphs did not reveal any $1$-Šoltés graph. We are only able to provide examples of large $\frac{1}{3}$-Šoltés graphs that are obtained by truncating certain cubic vertex-transitive graphs. This leads us to believe that no $1$-Šoltés graph other than $C_{11}$ exists.