The fundamental building blocks in band theory are band representations -bands whose infinitelynumbered Wannier functions are generated (by action of a space group) from a finite number of symmetric Wannier functions centered on a point in space. This work aims to simplify questions on a multi-rank band representation by splitting it into unit-rank bands, via the following crystallographic splitting theorem: being a rank-N band representation is equivalent to being splittable into a finite sum of bands indexed by {1, 2, . . . , N }, such that each band is spanned by a single, analytic Bloch function of k, and any symmetry in the space group acts by permuting {1, 2, . . . , N }. We prove this theorem for all band representations (of crystallographic space groups) whose Wannier functions transform in the integer-spin representation; in the half-integer-spin case, the only exceptions to the theorem exist for three-spatial-dimensional space groups with cubic point groups. Applying this theorem, we develop computationally efficient methods to determine whether a given energy band (of a tight-binding or Schrödinger-type Hamiltonian) is a band representation, and, if so, how to numerically construct the corresponding symmetric Wannier functions. Thus we prove that rotation-symmetric topological insulators in Wigner-Dyson class AI are fragile, meaning that the obstruction to symmetric Wannier functions can be removed by addition of band representations to the filled-band subspace. An implication of fragility is that its boundary states, while robustly covering the bulk energy gap in finite-rank tight-binding models, can be destabilized if the Hilbert space is expanded to include all symmetry-allowed representations. These fragile insulators have photonic analogs that we identify; in particular, we prove that an existing photonic crystal built by Yihao Yang et al. [Nature 565, 622 (2019)] is fragile topological with removable boundary states, which disproves a widespread perception of 'topologically-protected' boundary states in timereversal-invariant, gapped photonic/phononic crystals. As a final application of our theorem, we derive various symmetry obstructions on the Wannier functions of topological insulators; for certain space groups, these obstructions are proven to be equivalent to the nontrivial holonomy of Bloch functions.CONTENTS 46 G. Tightly-bound BRs and the existence of the symmetric tight-binding limit 47 1. G-vector bundles and tight-binding lattice models 47 2. BRs and tightly-bound BRs as G-vector bundles 48 3. Existence of symmetric tight-binding limit 48 H. Lemma for Zak phases of tightly-bound band representations 48 I. Proof of localization obstruction lemma 49 References 50